© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 4-RES-JB
    11. 6059
    Prev
    Next
    INDEX
    Explanations

    terms related to specialized markets or areas of focus, such as niches or specialties

    oai_token-act-pair · gpt-3.5-turbo

    words related to niche markets or specialties

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 4-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.4.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    4
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Fac
    -0.66
    âĢ¢âĢ¢âĢ¢âĢ¢
    -0.65
     largeDownload
    -0.64
    onz
    -0.64
    owitz
    -0.64
    shall
    -0.63
    agher
    -0.62
    Il
    -0.62
    endment
    -0.61
    actionDate
    -0.60
    POSITIVE LOGITS
     niche
    1.23
    odox
    0.78
     nic
    0.72
    markets
    0.66
     superpower
    0.66
     fry
    0.65
    holder
    0.64
     versa
    0.64
     locality
    0.62
    LLOW
    0.61
    Activations Density 0.008%

    No Known Activations